Individual differences in intrinsic radiosensitivity account for at least part of the observed variation in normal tissue and tumor responses to radiotherapy. However, normal tissue responses to radiation exposure are also influenced by biochemical signaling associated with cell growth, proliferation, differentiation status, to name a few, while tumors are influenced by these same signals as well as inappropriate signals due to mutation or other events. In the past, no single assay or collection of assays could capture all of this information in an informative manner so as to predict either the tumor or normal tissue response. We hypothesize that new methodologies that examine global gene expression are capable of capturing this information and that this information can be used to predict: normal tissue response; particularly late radiation sequelae; and the response of tumors, particularly Ioco-regional control, metastasis, and survival to radiotherapy. Furthermore, we submit that this information has clinical significance in that it can be generated in a timely fashion so that it can be used in making clinical decisions, may be used to identify clinical strategies for individuals, or perhaps identify new targets of opportunity for therapy. We will use microarray analysis to examine the underlying gene expression patterns of human skin cultures donated by patients whose in vitro SF2 values have previously been shown to correlate with the patient's clinical normal tissue response. We will attempt to link gene expression with the underlying sensitivity of these fibroblasts as well as to the clinical response of the normal tissue. Gene expression changes will be validated by secondary methodologies, at the protein level, and at the cellular or biochemical level. Once developed, a prognostic expression array will be tested against a series of head and neck cancer patients for validation. In parallel fashion, a series of head and neck tumors, from patients whose clinical fate is already known wil be examined for the underlying gene expression patterns that were responsible for their outcome. Where possible the expression of specific genes identified will be examined at the protein level by immunohistochemistry in the original tissue and in head and neck tumors in general by screening against a head and neck tumor tissue array. A prognostic expression array will be developed and tested against a series of head and neck patients to determine the prognostic power of the expression array.